An application of text mining to understand the productivity challenges in Singapore

The issue of low productivity growth and reliance on foreign labour are major concerns in Singapore. Traditionally, identifying productivity issues usually involves an expert´s opinion or market surveys. These methods might be biased or labour intensive. In this paper, we consider an alternative method, a data-driven approach using text mining. Various text mining techniques are applied on the data collected from 87 companies that participated in an operational excellence programme. The results indicated that even though the set of productivity issues for companies residing in Singapore are similar, the severity of the problems differs across different types of industries. Furthermore, we also found that companies from different industries tend to adopt different methods to solve their productivity issues.